Department of Radiology, Massachusetts General Hospital, Boston, Massachusetts, United States of America.
Harvard Medical School, Boston, Massachusetts, United States of America.
PLoS One. 2018 Oct 4;13(10):e0204155. doi: 10.1371/journal.pone.0204155. eCollection 2018.
Deep learning (DL) based solutions have been proposed for interpretation of several imaging modalities including radiography, CT, and MR. For chest radiographs, DL algorithms have found success in the evaluation of abnormalities such as lung nodules, pulmonary tuberculosis, cystic fibrosis, pneumoconiosis, and location of peripherally inserted central catheters. Chest radiography represents the most commonly performed radiological test for a multitude of non-emergent and emergent clinical indications. This study aims to assess accuracy of deep learning (DL) algorithm for detection of abnormalities on routine frontal chest radiographs (CXR), and assessment of stability or change in findings over serial radiographs.
We processed 874 de-identified frontal CXR from 724 adult patients (> 18 years) with DL (Qure AI). Scores and prediction statistics from DL were generated and recorded for the presence of pulmonary opacities, pleural effusions, hilar prominence, and enlarged cardiac silhouette. To establish a standard of reference (SOR), two thoracic radiologists assessed all CXR for these abnormalities. Four other radiologists (test radiologists), unaware of SOR and DL findings, independently assessed the presence of radiographic abnormalities. A total 724 radiographs were assessed for detection of findings. A subset of 150 radiographs with follow up examinations was used to asses change over time. Data were analyzed with receiver operating characteristics analyses and post-hoc power analysis.
About 42% (305/ 724) CXR had no findings according to SOR; single and multiple abnormalities were seen in 23% (168/724) and 35% (251/724) of CXR. There was no statistical difference between DL and SOR for all abnormalities (p = 0.2-0.8). The area under the curve (AUC) for DL and test radiologists ranged between 0.837-0.929 and 0.693-0.923, respectively. DL had lowest AUC (0.758) for assessing changes in pulmonary opacities over follow up CXR. Presence of chest wall implanted devices negatively affected the accuracy of DL algorithm for evaluation of pulmonary and hilar abnormalities.
DL algorithm can aid in interpretation of CXR findings and their stability over follow up CXR. However, in its present version, it is unlikely to replace radiologists due to its limited specificity for categorizing specific findings.
深度学习（DL）解决方案已被提出用于解释多种成像方式，包括放射摄影、CT 和 MR。对于胸部 X 光片，DL 算法已在评估肺结节、肺结核、肺囊性纤维化、尘肺和外周插入中心导管位置等异常方面取得成功。胸部 X 光片是最常用于多种非紧急和紧急临床适应症的最常见放射学检查。本研究旨在评估深度学习（DL）算法对常规正位胸部 X 光片（CXR）异常检测的准确性，并评估连续 X 光片上发现的稳定性或变化。
我们使用 DL（Qure AI）处理了 724 名成年患者（> 18 岁）的 874 张去识别正位 CXR。为了建立标准参考（SOR），两位胸部放射科医生评估了所有 CXR 的肺部混浊、胸腔积液、肺门突出和心脏轮廓增大的存在情况。来自 DL 的评分和预测统计数据为肺实质混浊、胸腔积液、肺门突出和心脏轮廓增大的存在情况生成并记录。为了确定标准参考（SOR），两位胸部放射科医生评估了所有 CXR 中这些异常的存在情况。其他四位放射科医生（测试放射科医生），不知道 SOR 和 DL 发现，独立评估了放射学异常的存在情况。共有 724 张 X 光片用于检测发现。使用具有随访检查的 150 张 X 光片子集来评估随时间的变化。使用接收器工作特性分析和事后功效分析进行数据分析。
根据 SOR，约 42%（305/724）的 CXR 没有发现；23%（168/724）和 35%（251/724）的 CXR 存在单一和多种异常。DL 和 SOR 之间的所有异常差异无统计学意义（p = 0.2-0.8）。DL 和测试放射科医生的曲线下面积（AUC）分别在 0.837-0.929 和 0.693-0.923 之间。DL 对随访 CXR 中肺实质混浊变化的评估具有最低的 AUC（0.758）。胸部植入设备的存在会降低 DL 算法评估肺和肺门异常的准确性。
DL 算法可辅助解释 CXR 结果及其随访 CXR 中的稳定性。然而，在其当前版本中，由于其对特定发现进行分类的特异性有限，因此不太可能替代放射科医生。
